基于分形分布的多重指示克里格估值方法及实例应用

Multiple indicator Kriging estimation method and a case study based on fractal distribution

地质统计学方法又称克里格法,是资源量估算的主流方法,当矿化元素品位呈混合总体分布且在空间上无法划分出有效的估算域时,宜采用多重指示克里格(MIK)方法。传统MIK估值一般选择十分位数作为指示阈值。但是该方法阈值的地质意义不明确,最低阈值和最高阈值内样品数据过少时,可能影响估值准确度,且建模过程繁琐。研究结果表明,矿化元素品位具有分形分布特征,可以用分形方法进行矿化混合总体筛分来确定MIK的指示阈值。本文以刚果(金)某铜钴矿床资源量估算为例,对比了分形分布法、概率图法和十分位法的MIK估值。研究结果显示,分形分布法MIK估值结果与十分位法基本一致,较传统十分位法建模过程简化,且有更明确的地质意义,作为传统十分位法MIK的改进方法具有推广应用价值。

Geostatistical method,also known as Kriging method,is the mainstream method of resource estimation.When the grade of mineralized elements presents mixed population distribution and no effective estimation domain can be spatially subdivided,the multiple indicator Kriging(MIK)method should be employed.Traditional MIK estimation generally chooses decile as indicator thresholds.However,the geological significance of the thresholds using this method is not clear,the accuracy of estimation may be affected when the sample data within the minimum and maximum thresholds are too few,and the modeling process is tedious.The results show that the grade of mineralized elements has the characteristics of fractal distribution,and the indicator thresholds of MIK can be determined by using the fractal method to decomposing the mineralized mixture population.Taking the resource estimation of a copper and cobalt deposit in DRC as an example,this paper compares the MIK estimation of the fractal distribution method,the probability graph method and the decile method.The results show that the MIK estimation based on the fractal distribution is basically consistent with the decile method.Moreover,the fractal distribution method has simpler modeling process and clearer geological significance compared with the traditional decile method.It has a great value of popularization and application as an improvement of the traditional decile MIK method.